The study, published in Cancer Causes & Control, drew on the Surveillance, Epidemiology, and End Results (SEER) 12 registry program, which covers approximately twelve percent of the U.S. population and offers continuous population-based incidence data stretching back to 1992. Researchers from Baylor College of Medicine identified 68,023 cases of hepatocellular carcinoma diagnosed between January 1, 1992 and December 31, 2022 among adults aged twenty and older. Rather than simply tracking raw incidence rates, the team focused on incidence rate ratios, or IRRs, which compare the age-adjusted rate in each demographic group against a reference group, women in the sex analysis and non-Hispanic White individuals in the race and ethnicity analysis. This approach isolates relative disparities from absolute incidence, revealing whether the distance between groups is widening or narrowing even when all rates are rising together.

The headline trend is that hepatocellular carcinoma in the United States has passed through an inflection point. Annual case counts climbed from 840 in 1992 to 2,846 in 2022, an increase of nearly 239 percent, and the overall age-adjusted rate rose from 4.27 to 7.84 per 100,000 person-years over the same period. Joinpoint regression, a statistical technique that identifies inflection points in trend data using Monte Carlo permutation testing, showed that incidence rose by 4.72 percent annually from 1992 to 2009, plateaued between 2009 and 2015, and then declined by 3.41 percent per year from 2015 to 2022. This late reversal is significant because earlier forecasts, notably a 2016 projection based on data through 2012, had anticipated continued increases through 2030. The new analysis demonstrates that the anticipated turning point has in fact materialized.

Sex differences remain the most durable feature of liver cancer epidemiology. Men were diagnosed at 12.24 per 100,000 in 2022 compared with 3.95 per 100,000 among women, and the male-to-female incidence rate ratio never dropped below three. Yet the ratio itself declined from 3.29 in 1992 to 3.11 in 2022. The dynamics behind that shift are revealing: from 1992 to 2012 the male-to-female IRR actually increased slightly, by about half a percent per year, before falling by 1.64 percent annually thereafter. The narrowing was driven by steeper declines in male incidence after 2012 rather than any surge in female rates, and the effect was most pronounced among adults under fifty, where the IRR dropped by roughly six percent per year after 2012, though the authors caution that the confidence interval on that age-stratified estimate was wide and warrants cautious interpretation.

The most dramatic change, however, unfolded along racial and ethnic lines. In 1992, non-Hispanic Asian/Pacific Islander individuals were diagnosed with hepatocellular carcinoma at nearly six times the rate of non-Hispanic White individuals, an IRR of 5.86. By 2022 that ratio had fallen to 1.87, an average annual decline of 3.93 percent across the study period. This extraordinary convergence reflects two simultaneous movements: incidence among Asian/Pacific Islander Americans actually declined over the three decades, from about 15.2 to 10.2 per 100,000, while incidence among non-Hispanic White individuals rose from 2.6 to 5.5 per 100,000. The declining IRR therefore captures both good news in one population and rising burden in another, a nuance the authors emphasize repeatedly.

The other relative disparities narrowed more modestly. The non-Hispanic Black-to-non-Hispanic White IRR fell from 2.08 to 1.44, with the most significant decline occurring between 2005 and 2022. The Hispanic-to-non-Hispanic White IRR barely moved, easing from 2.62 to 2.40, an average annual decrease of only 0.57 percent. Across the full study period, incidence remained highest among Asian/Pacific Islander individuals at 14.27 per 100,000, followed by Hispanic individuals at 12.12, non-Hispanic Black individuals at 9.30, and non-Hispanic White individuals at 4.37. Age-stratified analyses added further texture: among adults under fifty, the Black-to-White IRR declined linearly throughout the period, whereas among older adults it first rose by 1.61 percent annually through 2005 before declining by nearly two percent annually thereafter.

What lies behind these shifting ratios? The study cannot establish causation, since SEER registries do not record individual risk factors, but the temporal patterns align strikingly with known changes in the etiology of liver cancer. Hepatitis B virus has long been a dominant driver of hepatocellular carcinoma in Asian and Pacific Islander populations, and the United States added universal infant HBV vaccination to its immunization schedule in 1991, with particularly high uptake among post-1991 birth cohorts and among Asian American children. Because preventing chronic HBV infection in infancy reduces cancer risk only after a long latency, the timing of declining incidence among younger adults in these historically high-prevalence populations is temporally compatible with a vaccination effect, though the authors stress this remains hypothesis-generating rather than proof.

Hepatitis C tells a complementary story. Direct-acting antivirals, which cure HCV infection and thereby interrupt the progression to cirrhosis and cancer, became available in the United States in late 2013 and expanded rapidly through the latter half of the 2010s. The post-2015 decline in overall hepatocellular carcinoma incidence follows this therapeutic revolution closely enough to suggest a genuine contribution, but the variable latency between viral eradication and changes in cancer risk means the association must be interpreted cautiously. Meanwhile, a countervailing force is gathering strength: metabolic dysfunction-associated steatotic liver disease, or MASLD, and diabetes are becoming increasingly dominant drivers of new liver cancer cases, and population-level alcohol consumption and alcohol-associated liver disease mortality have both risen in recent decades. These trends do not explain the recent overall decline and may instead represent a competing influence shaping which groups continue to bear the burden.

The sex-specific narrowing may be an early signal of the metabolic transition. Accumulating evidence suggests that women with diabetes and MASLD may experience a disproportionately greater relative increase in liver-related risk compared with men, and as the prevalence of metabolic disease rises, particularly among younger and middle-aged women, the historical male predominance in liver cancer may continue to attenuate. This would represent a fundamental shift in the epidemiology of a cancer whose sex disparity has been attributed in part to higher rates of viral hepatitis and heavy alcohol use among men.

The study’s authors are careful about what these findings do and do not mean. The narrowing of incidence rate ratios signals attenuation of relative disparities, not their elimination; absolute incidence in 2022 remained substantially higher among men than women and among Asian/Pacific Islander, Hispanic, and Black individuals than among non-Hispanic White individuals. Limitations include the absence of individual-level data on risk factors, the possibility that evolving diagnostic practices and surveillance influenced case ascertainment, potential misclassification of race and ethnicity, and geographic coverage limited to roughly thirteen percent of the U.S. population, excluding regions such as Texas that appear only in newer SEER releases. The team selected SEER 12 over the broader SEER 21 database precisely because it extends back to 1992, enabling a genuinely long-term view. Still, the message for public health is clear: vaccination, antiviral therapy, and early detection have visibly reshaped who gets liver cancer in America, but continued efforts to prevent, detect, and manage both viral and metabolic liver disease are needed to close the remaining gaps.

Subject of Research: Temporal trends in sex and racial/ethnic disparities in hepatocellular carcinoma incidence in the United States from 1992 to 2022

Article Title: Temporal changes in sex and race/ethnicity incidence rate ratios for hepatocellular carcinoma in the United States, 1992–2022

Article References: Temporal changes in sex and race/ethnicity incidence rate ratios for hepatocellular carcinoma in the United States, 1992–2022. (n.d.). https://doi.org/10.1007/s10552-026-02238-w

Image Credits: AI Generated

DOI: 10.1007/s10552-026-02238-w

Keywords: hepatocellular carcinoma, liver cancer, incidence rate ratios, health disparities, SEER, hepatitis B, hepatitis C, MASLD, sex differences, race and ethnicity, cancer epidemiology, Cancer Causes & Control

The study, led by Soo Youn Bae of Seoul St. Mary’s Hospital at The Catholic University of Korea and colleagues, examined 696,960 women diagnosed with Stage I–III breast cancer between 1975 and 2021, drawing on the SEER 8 registries, the longest-running and most rigorously maintained cancer surveillance system in the United States. Rather than simply comparing average ages at the beginning and end of the period, the researchers modeled annual trajectories of diagnostic age using weighted ordinary least squares regression, estimating the rate of change—captured as a slope coefficient, β1—for each calendar year across every subgroup. The approach allowed them to detect not just whether diagnostic ages rose or fell, but how fast, at which cancer stage, and for whom.

The headline finding is a profound divergence along socioeconomic lines. Women living in the highest-income census tracts—the top quartile of neighborhood income—experienced significant, sustained increases in their age at diagnosis, consistent with the overall national trend toward later diagnosis. Women in the lowest-income quartile, by contrast, showed stagnant or even declining diagnostic ages. The pattern was particularly stark for Stage II disease: in the lowest-income group, the annual rate of change in diagnostic age was slightly negative, at β1 = −0.028 years per year, meaning that over the 46-year window, the average age at diagnosis for low-income women with Stage II breast cancer effectively moved backward while their wealthier counterparts aged into their diagnoses.

Geography told a parallel story. Women in rural communities, classified using the rural–urban continuum codes that rank counties from most metropolitan to most remote, similarly failed to participate in the national shift toward later diagnosis. Their diagnostic ages remained flat or drifted downward over the study period. The implication is uncomfortable: the demographic transition that has pushed breast cancer into later life for much of the population has simply bypassed the poorest and most geographically isolated women, whose tumors continue to be found at younger ages—and, given the established link between younger age at diagnosis for these groups and later-stage presentation, often at more dangerous points in the disease course.

The racial analysis added a second layer of complexity. Black women in the cohort showed the steepest annual increases in diagnostic age across all tumor stages, a trend that outpaced every other racial group. Yet despite this rapid upward trajectory, Black women remained younger at diagnosis than White women throughout the entire 46-year span. In other words, the fastest improvement in trajectory was not enough to close a persistent gap in the level. The finding captures a well-known paradox in breast cancer epidemiology: Black women are disproportionately diagnosed at younger ages and with more aggressive tumor subtypes, including higher rates of triple-negative and other hormone receptor–negative disease, while simultaneously facing barriers to timely screening and follow-up that delay detection within any given age band.

Why would diagnostic age rise for some groups and stall for others? The authors point to the intertwined machinery of screening access, reproductive and hormonal trends, and health care delivery. The national rise in diagnostic age partly reflects the aging of the population and the widespread adoption of screening mammography, which tends to detect cancers in older women earlier and more often. Mammography uptake, however, has never been evenly distributed. Studies spanning decades have documented lower screening rates among low-income women, rural residents, and some racial and ethnic minority groups, along with longer intervals between abnormal findings and diagnostic resolution. When screening is inconsistent, cancers are more likely to be detected symptomatically—and in groups with higher baseline risks of early-onset disease, that symptomatic detection skews young.

Hormone and reproductive factors plausibly deepen the divide. The rise and fall of menopausal hormone therapy—sharply curtailed after the Women’s Health Initiative reported increased breast cancer risk with combined estrogen plus progestin in 2002—altered incidence patterns, particularly among older, more affluent women who were most likely to use these therapies. Trends toward later childbearing, lower parity, and higher rates of obesity have reshaped risk profiles in ways that differ across socioeconomic strata. Higher-income women have, on balance, experienced risk-factor shifts associated with later-onset disease, while populations facing overlapping disadvantages have carried a heavier burden of early-onset, biologically aggressive tumors. The new trajectory data suggest these two forces have been quietly pulling the age of diagnosis apart for decades.

The technical rigor of the analysis lends weight to its conclusions. By stratifying simultaneously by stage, race, census tract income quartile, and rural–urban continuum code, and by applying weighted regression to annual mean diagnostic ages, the researchers could distinguish genuine temporal trends from artifacts of changing case mix. Stage-specific stratification matters because screening tends to shift the detected-stage distribution: rising diagnostic ages in Stage I disease can reflect early detection in older women, whereas declining or flat diagnostic ages in Stage II–III disease signal that some populations are not being caught by the early-detection net at all. The negative slope for low-income Stage II patients is precisely the signature one would expect if early-onset disease continues to dominate in a population that screening programs have failed to reach.

The policy implications are difficult to ignore. Current screening guidelines in the United States are built around age thresholds—typically recommending that average-risk women begin mammography in their 40s or 50s—implicitly assuming that breast cancer risk rises with age in a broadly uniform way. This study challenges that assumption by showing that the age structure of risk has diverged across social strata. A one-size-fits-all age-based policy, the authors argue, entrenches inequity: it calibrates optimally for high-income metropolitan women, whose diagnostic ages are rising, while under-serving low-income and rural women, whose disease continues to present earlier. Equitable, subgroup-specific screening strategies—whether through risk-adapted starting ages, enhanced outreach in underserved communities, mobile mammography in rural areas, or patient-navigation programs to shorten diagnostic delays—follow directly from the data.

The study also carries a broader lesson about how aggregate statistics can conceal inequality. Had the researchers stopped at the national average, the story would have been one of gradual progress: women, on the whole, are developing breast cancer later, and later diagnosis within a screened population generally correlates with better outcomes. Only by decomposing the trend did the deeper reality emerge—that the benefits of a half-century of progress in cancer detection and care have been distributed unevenly along lines of income, geography, and race. The two-year rise in mean diagnostic age is real, but it is an average of two very different worlds: one where women age into screenings and early detections, and another where cancer arrives early, often between screenings, and frequently at a stage that screening was supposed to prevent. Closing that diagnostic age gap, the authors conclude, is now one of the clearest quantitative targets for achieving equity in breast cancer control.

Subject of Research: Temporal trends in age at breast cancer diagnosis across socioeconomic, racial, and geographic subgroups in the United States from 1975 to 2021

Article Title: Temporal trajectories of age at breast cancer diagnosis by socioeconomic and geographic factors: a 1975–2021 SEER analysis

Article References: Bae, S. Y., Kim, C. W., Chin, J., Lee, J. A., Kim, D., Lee, Y. J., Yoon, C. I., & Park, W.-C. (2026). Temporal trajectories of age at breast cancer diagnosis by socioeconomic and geographic factors: a 1975–2021 SEER analysis. Cancer Causes & Control, 37(10), Article 160. https://doi.org/10.1007/s10552-026-02247-9

Image Credits: AI Generated

DOI: 10.1007/s10552-026-02247-9

Keywords: breast cancer, age at diagnosis, health disparities, socioeconomic factors, SEER, cancer screening, rural health, epidemiology, health equity, mammography, social determinants of health, cancer surveillance